PROJECT SUMMARY This project addresses a basic research priority of the NEI, which is to understand how retinal networks process visual images. The mammalian retina contains about 20 different types of retinal ganglion cell each having characteristic receptive field properties. Most ganglion cell types are image- forming in that they comprise an array of essentially identical neurons that completely sample the visual field. Our approach has been to target specific, anatomically and physiologically identified ganglion cells, and perform quantitative analyses of the light-evoked inputs in order to gain insights into the connectivity and neural pathways that mediate physiological responses. In this proposal we will address three aims. Aim 1 proposes to test the hypothesis that selective expression of AMPA and kainate receptors in the outer plexiform layer contribute to temporal tuning in the ganglion cells. Aim 2 proposes to examine the role of NMDA receptors in generating the receptive field properties of direction-selective ganglion cells, and ON brisk-sustained ganglion cells. We will use subunit-specific antagonists to probe differences in NMDA receptor function in the ON and OFF pathways. Aim 3, in collaboration with Robert Smith at University of Pennsylvania, will use realistic computational models to predict the synaptic conductances, and estimate the magnitude of voltage-clamp errors in the ganglion cell recordings.